GEP Book   Home   News   Author   Q&A   Tutorials   Downloads   GEP Biblio   Contacts   Visit Gepsoft

C. FERREIRA Invited Tutorial Presented at WSC6, 2001 Gene Expression Programming in Problem Solving Function Finding   We have already seen how GEP can be used to do symbolic regression in the simple example of section 2.4. However, despite the fact that the target function contained simple numerical constants (3, 2, and 1), there was no explicit facility to generate them: the algorithm created them on its own. In this section, I will show how GEP solves the problem of explicit constant creation to do symbolic regression. Furthermore, I will also show that the explicit use of constants is, in fact, better avoided, for GEP produces better results when the numerical constants are not explicitly used and the system is left to invent new ways of representing them. Home | Contents | Previous | Next